------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_logs/figure5.log
  log type:  text
 opened on:  23 Jun 2022, 11:42:48

. 
. ** panel a 
. if 1 {
.         
.         use "$Rep_smokelabor/1_build/regdata/agegrp_county_quarter.dta", clear 
.         
.         foreach grp in A04 A05 A06 A07 A08 {
  2.         preserve    
  3.                 
.                 keep if agegrp=="`grp'"
  4.                 tsset fe_countyqtroy rfrnc_yr
  5.                 
.                 * first diff: y(t) minus y(t-1)
.                 foreach v of varlist pc_qwi_payroll {
  6.                         gen d_`v'=`v'-L1.`v'
  7.                 }
  8.                 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros)
  9.         
.         restore         
 10.         }
(721,504 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(18,866 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.9246e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    23.20
                                                      Prob > F      =   0.0000
Total (centered) SS     =  3.91491e+10                Centered R2   =  -0.0437
Total (uncentered) SS   =  3.91491e+10                Uncentered R2 =  -0.0437
Residual SS             =  4.08581e+10                Root MSE      =    742.6

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -134.3182   27.88618    -4.82   0.000    -189.0134   -79.62296
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             49.246
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2988.712
                         (Kleibergen-Paap rk Wald F statistic):         67.813
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(721,504 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(18,866 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.8343e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    23.18
                                                      Prob > F      =   0.0000
Total (centered) SS     =  5.42253e+10                Centered R2   =  -0.0588
Total (uncentered) SS   =  5.42253e+10                Uncentered R2 =  -0.0588
Residual SS             =  5.74151e+10                Root MSE      =    880.3

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -183.5145   38.11429    -4.81   0.000    -258.2708   -108.7581
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             49.988
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2905.289
                         (Kleibergen-Paap rk Wald F statistic):         69.343
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(721,504 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(18,866 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.9192e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    23.93
                                                      Prob > F      =   0.0000
Total (centered) SS     =  4.09931e+10                Centered R2   =  -0.1203
Total (uncentered) SS   =  4.09931e+10                Uncentered R2 =  -0.1203
Residual SS             =  4.59265e+10                Root MSE      =    787.3

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -216.8324   44.32462    -4.89   0.000    -303.7695   -129.8953
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.268
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2757.817
                         (Kleibergen-Paap rk Wald F statistic):         71.256
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(721,504 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(18,866 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.6897e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    27.72
                                                      Prob > F      =   0.0000
Total (centered) SS     =  2.48473e+10                Centered R2   =  -0.1322
Total (uncentered) SS   =  2.48473e+10                Uncentered R2 =  -0.1322
Residual SS             =  2.81325e+10                Root MSE      =    616.2

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -176.6969   33.56195    -5.26   0.000    -242.5244   -110.8693
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             53.842
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2922.198
                         (Kleibergen-Paap rk Wald F statistic):         72.803
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(721,504 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(18,866 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.9304e+09)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    25.19
                                                      Prob > F      =   0.0000
Total (centered) SS     =   1550465616                Centered R2   =  -0.1089
Total (uncentered) SS   =   1550465616                Uncentered R2 =  -0.1089
Residual SS             =   1719234536                Root MSE      =    152.3

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -38.59237   7.689181    -5.02   0.000    -53.67373   -23.51102
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.168
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2987.939
                         (Kleibergen-Paap rk Wald F statistic):         69.096
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.         
.         ** plot 
.         import excel using "$Rep_smokelabor/2_analysis/output_figures/figure5.xlsx", sheet("age") clear first
(11 vars, 5 obs)
.         
.         label define agegrp_lb 1 "age 25-34" 2 "age 35-44" 3 "age 45-54" 4 "age 55-64" 5 "age 65+", replace
.         label values agegrp agegrp_lb
.         
.         gen pct_inc=b_inc*100/m_inc
.         
.         tw      bar pct_inc agegrp, yaxis(1) col(teal%30) lcol(white) lw(0.001) barw(0.4) || ///
>                 rcap l_inc u_inc agegrp, col(gs9) lp(dash) yaxis(2) || ///
>                 scatter b_inc agegrp, col(blue) yaxis(2) ///
>                 ylab(-350(70)0, nogrid axis(2)) ylab(-3(0.6)0, axis(1) nogrid) ///
>                 xtitle("") ytitle("") ///
>                 yline(0, lcol(black) lp(dot) lw(0.6)) ///
>                 xlab(1 2 3 4 5,valuelabel labsize(small)) ///
>                 legend(col(2) size(small) order(1 "% change (left axis)" 3 "\$ change (right axis)")) ///
>                 graphregion(col(white))
.         gr export "$Rep_smokelabor/2_analysis/output_figures/figure5_a.pdf", replace
(file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_figures/figure5_a.pdf written in PDF format)
.                                                                 
. }

. 
. ** panel b
. if 1 {
.         
.         use "$Rep_smokelabor/1_build/regdata/naics2_county_quarter.dta", clear 
.         
.         forv grp =1/20 {
  2.         preserve    
  3.                 
.                 keep if g_industry==`grp'
  4.                 tsset fe_countyqtroy rfrnc_yr
  5.                 
.                 * 1-year change: y(t) minus y(t-1)
.                 foreach v of varlist pc_qwi_payroll {
  6.                         gen d_`v'=`v'-L1.`v'
  7.                 }
  8.                 
.                 ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip fe_stqtros)
  9.         
.         restore         
 10.         }
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(23,623 missing values generated)
(dropped 209 singleton observations)
(sum of wgt is     1.3826e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1675               Number of obs =    73030
Number of clusters (fe_stqtros) =   2476              F(  1,  1674) =     0.32
                                                      Prob > F      =   0.5733
Total (centered) SS     =  4128449.891                Centered R2   =   0.0003
Total (uncentered) SS   =  4128449.891                Uncentered R2 =   0.0003
Residual SS             =  4127151.766                Root MSE      =     7.55

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |     .09278   .1646966     0.56   0.573     -.230253     .415813
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.288
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2834.245
                         (Kleibergen-Paap rk Wald F statistic):         71.643
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6621        6621           0    *|
        fe_styr |       620           0         620     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(56,731 missing values generated)
(dropped 258 singleton observations)
(sum of wgt is     1.3248e+10)
(MWFE estimator converged in 9 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1496               Number of obs =    63168
Number of clusters (fe_stqtros) =   2457              F(  1,  1495) =        .
                                                      Prob > F      =        .
Total (centered) SS     =  109146080.6                Centered R2   =  -0.0002
Total (uncentered) SS   =  109146080.6                Uncentered R2 =  -0.0002
Residual SS             =  109172981.4                Root MSE      =    41.78

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -.3618112          .        .       .            .           .
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             50.315
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2466.454
                         (Kleibergen-Paap rk Wald F statistic):         69.153
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Warning: estimated covariance matrix of moment conditions not of full rank.
         overidentification statistic not reported, and standard errors and
         model tests should be interpreted with caution.
Possible causes:
         number of clusters insufficient to calculate robust covariance matrix
         singleton dummy variable (dummy with one 1 and N-1 0s or vice versa)
partial option may address problem.
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      5886        5886           0    *|
        fe_styr |       616           0         616     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(28,607 missing values generated)
(dropped 196 singleton observations)
(sum of wgt is     1.3852e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1649               Number of obs =    72284
Number of clusters (fe_stqtros) =   2520              F(  1,  1648) =    27.83
                                                      Prob > F      =   0.0000
Total (centered) SS     =   62959925.7                Centered R2   =  -0.0037
Total (uncentered) SS   =   62959925.7                Uncentered R2 =  -0.0037
Residual SS             =  63193031.07                Root MSE      =     29.7

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1.454814   .2757937    -5.28   0.000    -1.995757   -.9138713
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.176
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2787.194
                         (Kleibergen-Paap rk Wald F statistic):         71.485
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6527        6527           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(19,535 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74687
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    21.00
                                                      Prob > F      =   0.0000
Total (centered) SS     =  188195725.9                Centered R2   =  -0.0114
Total (uncentered) SS   =  188195725.9                Uncentered R2 =  -0.0114
Residual SS             =  190350160.9                Root MSE      =     50.7

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -4.054733   .8848298    -4.58   0.000    -5.790215   -2.319251
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.436
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2882.502
                         (Kleibergen-Paap rk Wald F statistic):         71.832
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(22,696 missing values generated)
(dropped 190 singleton observations)
(sum of wgt is     1.3906e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1680               Number of obs =    74393
Number of clusters (fe_stqtros) =   2520              F(  1,  1679) =    26.06
                                                      Prob > F      =   0.0000
Total (centered) SS     =  500493651.1                Centered R2   =  -0.0461
Total (uncentered) SS   =  500493651.1                Uncentered R2 =  -0.0461
Residual SS             =  523588617.4                Root MSE      =    84.25

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -14.65219   2.870024    -5.11   0.000     -20.2814   -9.022993
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.416
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2870.563
                         (Kleibergen-Paap rk Wald F statistic):         71.800
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6654        6654           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(20,986 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.3906e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74571
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    25.53
                                                      Prob > F      =   0.0000
Total (centered) SS     =  83991941.34                Centered R2   =  -0.0559
Total (uncentered) SS   =  83991941.34                Uncentered R2 =  -0.0559
Residual SS             =   88686146.5                Root MSE      =    34.63

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -6.521224   1.290727    -5.05   0.000    -9.052822   -3.989626
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.431
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2877.864
                         (Kleibergen-Paap rk Wald F statistic):         71.824
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(19,140 missing values generated)
(dropped 188 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1684               Number of obs =    74689
Number of clusters (fe_stqtros) =   2520              F(  1,  1683) =    27.50
                                                      Prob > F      =   0.0000
Total (centered) SS     =  117331482.3                Centered R2   =  -0.0265
Total (uncentered) SS   =  117331482.3                Uncentered R2 =  -0.0265
Residual SS             =  120437816.2                Root MSE      =    40.33

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -5.622483   1.072205    -5.24   0.000    -7.725477   -3.519488
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.435
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2882.554
                         (Kleibergen-Paap rk Wald F statistic):         71.831
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6672        6672           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(21,193 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.3906e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1684               Number of obs =    74539
Number of clusters (fe_stqtros) =   2520              F(  1,  1683) =    20.14
                                                      Prob > F      =   0.0000
Total (centered) SS     =  104156607.1                Centered R2   =  -0.0162
Total (uncentered) SS   =  104156607.1                Uncentered R2 =  -0.0162
Residual SS             =  105840546.3                Root MSE      =    37.84

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -3.879716   .8645428    -4.49   0.000    -5.575408   -2.184024
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.427
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2876.810
                         (Kleibergen-Paap rk Wald F statistic):         71.817
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6672        6672           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(22,954 missing values generated)
(dropped 190 singleton observations)
(sum of wgt is     1.3903e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1677               Number of obs =    74161
Number of clusters (fe_stqtros) =   2520              F(  1,  1676) =     2.65
                                                      Prob > F      =   0.1040
Total (centered) SS     =  397180672.1                Centered R2   =  -0.0031
Total (uncentered) SS   =  397180672.1                Uncentered R2 =  -0.0031
Residual SS             =  398420088.6                Root MSE      =    73.61

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1.723377   1.059585    -1.63   0.104    -3.801627    .3548729
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.408
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2861.475
                         (Kleibergen-Paap rk Wald F statistic):         71.791
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6644        6644           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(19,888 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1683               Number of obs =    74644
Number of clusters (fe_stqtros) =   2520              F(  1,  1682) =     9.53
                                                      Prob > F      =   0.0021
Total (centered) SS     =   2113270615                Centered R2   =  -0.0085
Total (uncentered) SS   =   2113270615                Uncentered R2 =  -0.0085
Residual SS             =   2131254142                Root MSE      =    169.7

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -13.74032    4.45032    -3.09   0.002    -22.46907   -5.011576
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.432
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2880.672
                         (Kleibergen-Paap rk Wald F statistic):         71.825
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6668        6668           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(26,213 missing values generated)
(dropped 196 singleton observations)
(sum of wgt is     1.3903e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1677               Number of obs =    74037
Number of clusters (fe_stqtros) =   2520              F(  1,  1676) =    24.76
                                                      Prob > F      =   0.0000
Total (centered) SS     =  9906955.892                Centered R2   =  -0.0421
Total (uncentered) SS   =  9906955.892                Uncentered R2 =  -0.0421
Residual SS             =  10324524.29                Root MSE      =    11.86

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -1.952502   .3924249    -4.98   0.000    -2.722196   -1.182807
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.392
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2856.454
                         (Kleibergen-Paap rk Wald F statistic):         71.763
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6640        6640           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(20,393 missing values generated)
(dropped 188 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74641
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    10.13
                                                      Prob > F      =   0.0015
Total (centered) SS     =   1100873141                Centered R2   =  -0.0052
Total (uncentered) SS   =   1100873141                Uncentered R2 =  -0.0052
Residual SS             =   1106633198                Root MSE      =    122.3

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -8.429825   2.648328    -3.18   0.001    -13.62419   -3.235463
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.433
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2880.694
                         (Kleibergen-Paap rk Wald F statistic):         71.828
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6675        6675           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(65,382 missing values generated)
(dropped 263 singleton observations)
(sum of wgt is     1.3717e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1524               Number of obs =    64099
Number of clusters (fe_stqtros) =   2520              F(  1,  1523) =    21.35
                                                      Prob > F      =   0.0000
Total (centered) SS     =  137234547.2                Centered R2   =  -0.0169
Total (uncentered) SS   =  137234547.2                Uncentered R2 =  -0.0169
Residual SS             =  139554369.6                Root MSE      =    46.89

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -4.906845   1.062068    -4.62   0.000    -6.990116   -2.823575
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             51.432
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2460.929
                         (Kleibergen-Paap rk Wald F statistic):         70.258
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      5984        5984           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(23,442 missing values generated)
(dropped 191 singleton observations)
(sum of wgt is     1.3906e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74474
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    24.69
                                                      Prob > F      =   0.0000
Total (centered) SS     =  82679383.11                Centered R2   =  -0.0355
Total (uncentered) SS   =  82679383.11                Uncentered R2 =  -0.0355
Residual SS             =  85610425.79                Root MSE      =    34.05

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -5.484294   1.103717    -4.97   0.000    -7.649095   -3.319493
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.425
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2874.084
                         (Kleibergen-Paap rk Wald F statistic):         71.815
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6671        6671           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(19,260 missing values generated)
(dropped 192 singleton observations)
(sum of wgt is     1.3897e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74664
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    20.22
                                                      Prob > F      =   0.0000
Total (centered) SS     =  225771390.6                Centered R2   =  -0.0103
Total (uncentered) SS   =  225771390.6                Uncentered R2 =  -0.0103
Residual SS             =    228086805                Root MSE      =    55.51

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -4.888313   1.087118    -4.50   0.000    -7.020557   -2.756069
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.438
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2885.581
                         (Kleibergen-Paap rk Wald F statistic):         71.841
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6671        6671           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(19,747 missing values generated)
(dropped 190 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1684               Number of obs =    74673
Number of clusters (fe_stqtros) =   2520              F(  1,  1683) =    34.79
                                                      Prob > F      =   0.0000
Total (centered) SS     =  189927441.3                Centered R2   =  -0.1309
Total (uncentered) SS   =  189927441.3                Uncentered R2 =  -0.1309
Residual SS             =  214787247.8                Root MSE      =    53.86

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -16.37909   2.776827    -5.90   0.000    -21.82548   -10.93269
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.436
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2881.937
                         (Kleibergen-Paap rk Wald F statistic):         71.831
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6672        6672           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(28,405 missing values generated)
(dropped 199 singleton observations)
(sum of wgt is     1.3891e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1670               Number of obs =    73238
Number of clusters (fe_stqtros) =   2520              F(  1,  1669) =     0.32
                                                      Prob > F      =   0.5719
Total (centered) SS     =  578719739.2                Centered R2   =   0.0005
Total (uncentered) SS   =  578719739.2                Uncentered R2 =   0.0005
Residual SS             =    578403614                Root MSE      =    89.25

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -.7551458   1.335534    -0.57   0.572    -3.374643    1.864352
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.377
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2827.460
                         (Kleibergen-Paap rk Wald F statistic):         71.739
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6605        6605           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(19,756 missing values generated)
(dropped 189 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1684               Number of obs =    74670
Number of clusters (fe_stqtros) =   2520              F(  1,  1683) =    28.91
                                                      Prob > F      =   0.0000
Total (centered) SS     =  19430608.22                Centered R2   =  -0.0374
Total (uncentered) SS   =  19430608.22                Uncentered R2 =  -0.0374
Residual SS             =  20157094.24                Root MSE      =     16.5

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -2.790788   .5190629    -5.38   0.000    -3.808864   -1.772711
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.433
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2881.707
                         (Kleibergen-Paap rk Wald F statistic):         71.827
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6672        6672           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(20,038 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    31.67
                                                      Prob > F      =   0.0000
Total (centered) SS     =  31162380.22                Centered R2   =  -0.0140
Total (uncentered) SS   =  31162380.22                Uncentered R2 =  -0.0140
Residual SS             =  31597154.51                Root MSE      =    20.65

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |   -2.10784   .3745277    -5.63   0.000    -2.842429   -1.373251
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.437
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2883.998
                         (Kleibergen-Paap rk Wald F statistic):         71.833
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
(3,427,144 observations deleted)
       panel variable:  fe_countyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
(18,925 missing values generated)
(dropped 187 singleton observations)
(sum of wgt is     1.3907e+10)
(MWFE estimator converged in 8 iterations)

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity and clustering on countyfip and fe_stqtros

Number of clusters (countyfip) =   1685               Number of obs =    74725
Number of clusters (fe_stqtros) =   2520              F(  1,  1684) =    22.18
                                                      Prob > F      =   0.0000
Total (centered) SS     =  236970538.8                Centered R2   =  -0.0053
Total (uncentered) SS   =  236970538.8                Uncentered R2 =  -0.0053
Residual SS             =  238224563.5                Root MSE      =     56.7

------------------------------------------------------------------------------
             |               Robust
d_pc_qwi_p~l |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        pm25 |  -3.676927   .7807737    -4.71   0.000    -5.208316   -2.145538
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic):             52.437
                                                   Chi-sq(1) P-val =    0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):             2883.998
                         (Kleibergen-Paap rk Wald F statistic):         71.833
Stock-Yogo weak ID test critical values: 10% maximal IV size             16.38
                                         15% maximal IV size              8.96
                                         20% maximal IV size              6.66
                                         25% maximal IV size              5.53
Source: Stock-Yogo (2005).  Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):         0.000
                                                 (equation exactly identified)
------------------------------------------------------------------------------
Instrumented:         pm25
Excluded instruments: hms_deep
Partialled-out:       _cons
                      nb: total SS, model F and R2s are after partialling-out;
                          any small-sample adjustments include partialled-out
                          variables in regressor count K
------------------------------------------------------------------------------

Absorbed degrees of freedom:
--------------------------------------------------------+
    Absorbed FE | Categories  - Redundant  = Num. Coefs |
----------------+---------------------------------------|
 fe_countyqtroy |      6676        6676           0    *|
        fe_styr |       631           0         631     |
--------------------------------------------------------+
* = FE nested within cluster; treated as redundant for DoF computation
.         
.         egen fe_indcountyqtroy=group(industry countyfip rfrnc_qtroy)
.         tsset fe_indcountyqtroy rfrnc_yr
       panel variable:  fe_indcountyqtroy (strongly balanced)
        time variable:  rfrnc_yr, 2006 to 2019
                delta:  1 unit
.         
.         * first diff: y(t) minus y(t-1)
.         foreach v of varlist pc_qwi_payroll {
  2.                 gen d_`v'=`v'-L1.`v'
  3.         }
(516,914 missing values generated)
.         
.         compress g_industry
  (0 bytes saved)
.         local cmd ivreghdfe OUTCOMEVAR  (pm25=hms_deep) [aw=seer_pop] , a(fe_countyqtroy fe_styr) cluster(countyfip)
.         wyoung d_pc_qwi_payroll, cmd(`cmd') familyp(pm25) bootstraps(100) subgroup(g_industry) seed(20) cluster(countyfip) replace
Estimating family-wise adjusted p-values for 20 hypothesis tests
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==1, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==2, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==3, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==4, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==5, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==6, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==7, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==8, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==9, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==10, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==11, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==12, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==13, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==14, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==15, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==16, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==17, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==18, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==19, a(fe_countyqtroy fe_styr) cluster(countyfip)
    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==20, a(fe_countyqtroy fe_styr) cluster(countyfip)

     +-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
     |  k                                                                                                                       model            outcome   familyp   subgroup         coef      stderr           p   pwyoung       pbonf      psidak |
     |-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
  1. |  1    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==1, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          1       .09278   .10208229   .36354716       .74           1   .74219064 |
  2. |  2    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==2, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          2   -.36181122    .6269597   .56396657       .85           1   .80987485 |
  3. |  3    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==3, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          3   -1.4548144   .23483026   7.338e-10         0   8.806e-09   8.806e-09 |
  4. |  4    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==4, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          4   -4.0547328   .59006634   8.904e-12         0   1.158e-10   1.158e-10 |
  5. |  5    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==5, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          5   -14.652195    2.010022   4.767e-13         0   7.627e-12   7.628e-12 |
     |-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
  6. |  6    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==6, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          6   -6.5212241   .83142082   7.721e-15         0   1.467e-13   1.477e-13 |
  7. |  7    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==7, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          7   -5.6224827    .7504675   1.090e-13         0   1.853e-12   1.853e-12 |
  8. |  8    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==8, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          8   -3.8797158   .81435246   2.059e-06       .02   .00002059   .00002059 |
  9. |  9    ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==9, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25          9   -1.7233769   1.3197344   .19178234        .7   .76712938   .57330915 |
 10. | 10   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==10, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         10   -13.740325   4.5057499   .00232793        .4   .01163966   .01158559 |
     |-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
 11. | 11   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==11, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         11   -1.9525019   .35235509   3.481e-08       .01   3.829e-07   3.829e-07 |
 12. | 12   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==12, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         12   -8.4298246   2.4746928   .00067371        .3   .00404224   .00403543 |
 13. | 13   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==13, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         13   -4.9068452   1.2612533   .00010437       .12   .00083496   .00083466 |
 14. | 14   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==14, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         14   -5.4842942   .76712123   1.296e-12         0   1.944e-11   1.944e-11 |
 15. | 15   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==15, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         15    -4.888313   1.3779437    .0003995       .23   .00279649   .00279314 |
     |-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
 16. | 16   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==16, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         16   -16.379086   1.7775245   9.029e-20         0   1.806e-18           0 |
 17. | 17   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==17, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         17   -.75514581   1.4336422   .59844943       .85           1   .80987485 |
 18. | 18   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==18, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         18   -2.7907876   .39595269   2.633e-12         0   3.686e-11   3.686e-11 |
 19. | 19   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==19, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         19     -2.10784   .27682492   4.390e-14         0   7.902e-13   7.894e-13 |
 20. | 20   ivreghdfe d_pc_qwi_payroll (pm25=hms_deep) [aw=seer_pop]  if g_industry==20, a(fe_countyqtroy fe_styr) cluster(countyfip)   d_pc_qwi_payroll      pm25         20   -3.6769268   .80362007   5.099e-06       .02    .0000459   .00004589 |
     +-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
.         saveold "$Rep_smokelabor/2_analysis/output_figures/figure5_b_westfall_young.dta", replace 
(saving in Stata 13 format)
(FYI, saveold has options version(12) and version(11) that write files in older Stata formats)
(note: file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_figures/figure5_b_westfall_young.dta not found)
file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_figures/figure5_b_westfall_young.dta saved
.         
.         ** plot 
.         import excel using "$Rep_smokelabor/2_analysis/output_figures/figure5.xlsx", sheet("ind") clear first
(7 vars, 20 obs)
. 
.         label define industrygrp_lb 1 "Agriculture" 2 "Mining" 3 "Utility" 4 "Construction" 5 "Manufacturing" 6 "Wholesale" 7 "Retail" 8 "Transportation" 9 "Information" 10 "Finance" 11 "Real estate" 12 "Professional" 13 "Management" 14 "Admin" 15 "E
> ducation" 16 "Healthcare" 17 "Entertainment" 18 "Accommodation" 19 "Other services" 20 "Public admin" , replace
.         label values industrygrp industrygrp_lb
. 
.         gen pct_inc=b_inc*100/m_inc
.         gsort pct_inc
.         gen s_industrygrp=_n
.         
.         tw      bar pct_inc s_industrygrp,  horizontal yaxis(1) col(teal%30) lcol(white) lw(0.001) barw(0.4) || ///
>                 rcap l_inc u_inc s_industrygrp, horizontal col(gs9) lp(dash) xaxis(2)   || ///
>                 dot b_inc s_industrygrp, dotex(no) ndots(0) horizontal col(blue) xaxis(2) || ///
>                 dot b_inc s_industrygrp if inlist(industrygrp,1,2,9,10,12,13,15,17), dotex(no) ndots(0) horizontal col(white) msize(0.8) xaxis(2)  ///
>                 xline(0, lcol(teal*0.5)) ///
>                 ytitle("") ///
>                 xtitle("Coefficient") ///
>                 legend(col(2) size(small) keygap(0.5) region(lcol(white)) order(1 "% change (bottom axis)" 3 "# change (top axis)")) ///
>                 ylab(1(1)20, nogrid valuelabel angle(0) labs(vsmall)) ///
>                 xlab(-4(1)1, axis(1)) xlab(-24(6)6, axis(2)) ///
>                 xtitle("", axis(1)) xtitle("", axis(2)) ///
>                 graphregion(color(white)) ///
>                 ysize(5.5) yscale(reverse)
.         gr export "$Rep_smokelabor/2_analysis/output_figures/figure5_b.pdf", replace
(file C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_figures/figure5_b.pdf written in PDF format)
. }

. 
. log close
      name:  <unnamed>
       log:  C:\Users\ericzou\Dropbox\replicate_smokelabor/2_analysis/output_logs/figure5.log
  log type:  text
 closed on:  24 Jun 2022, 17:42:18
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
